What is account scoring?

How does account scoring work?

An account score adds up points for two kinds of evidence, fit and timing, and sorts the list by the total, so that the accounts most likely to buy now are worked first.

• Fit: how closely the company matches your ideal customer profile, on industry, size, location and technology. Fit changes slowly.
• Timing: signals that the company may buy now, such as hiring for the team your product serves, adopting a tool you integrate with, raising money or changing leaders. Timing can change from one week to the next.

Each criterion gets a weight, the weights add up to a score, and the score sorts the list. Teams then cut the sorted list into tiers: the top accounts get research and personal outreach, the middle ones a lighter sequence, the rest wait for a signal.

Fit

HighLow

• Wait for a signalA lighter sequence, and a watch on the account
• Work it nowResearch and personal outreach
• Leave outOutside the profile, and quiet
• Check the fit firstA signal at the wrong company is still the wrong company

No signal yetSignal now

Timing

A score is only as good as the criteria behind it. The useful question is not “how high is this score?” but “which evidence put it there?”

What signals go into an account score?

An account score can draw on four families of signals, from the slowest to the fastest:

Signal	Example	What it tells you
Firmographic fit	Software, 201–500 employees, headquartered in the US	Whether the company belongs in your market
Technographic fit	Runs a CRM you integrate with	Whether your product fits its stack
Hiring	Posted Account Executive roles this month	That a team is growing, and which one
Intent and engagement	Researches your category, visits your pricing page	That someone is looking now

When several independent signals point to the same company, they deserve more weight than any one of them alone. The TAM guide builds a whole market this way and calls it signal convergence.

What is rule-based scoring?

Rule-based scoring is a model whose rules are written by people, such as “2 points if the company has 201 to 500 employees, 1 point per sales role posted this month”. It is transparent, easy to change and the right place to start. Predictive scoring learns the weights from past wins instead: it needs enough closed deals to learn from, and its reasons are harder to read.

The example above applies a small rule-based model, written for a seller of a tool for sales teams:

Criterion	Points
Fit: 201–500 employees, the core of the ICP	2
Fit: 51–200 employees	1
Timing: each sales role posted in the last 30 days, up to 3	1

Applied to the four companies, it gives this order:

Company	Fit	Sales roles, last 30 days	Example score
Lucidity	1	2	3
Sprinto	2	0	2
Joveo	2	0	2
Quartic	1	0	1

Three things to notice. Lucidity comes first despite the weaker fit, because it is building the team the product serves. Joveo posted 198 jobs in the same 30 days and still earns no timing point: for this seller, the role matters, not the volume. And Sprinto and Joveo tie, because nothing in the model separates them; a tie is a cue to add a criterion or let a person decide. The score orders outreach for one seller under one model. It says nothing about the quality of these companies.

Check yourselfYou sell a tool for sales teams, and the core of your ICP is software companies with 201–500 employees. Which account should a fit-plus-timing model put first?

• A 300-person software company that posted 40 engineering jobs and no sales role this month
• A 250-person software company that posted three Account Executive roles this month
• A 6,000-person bank that posted ten sales roles this month
• A 120-person software company with no job posted at all

Show the answer key

1. ✕A 300-person software company that posted 40 engineering jobs and no sales role this monthStrong fit, but its hiring says nothing about the team your tool serves: volume alone is not a timing signal.
2. ✓A 250-person software company that posted three Account Executive roles this monthCore fit and the right signal: it is growing the team that would use your tool.
3. ✕A 6,000-person bank that posted ten sales roles this monthThe signal is right, but the bank is outside the ICP: fit decides who is on the list, timing only orders it.
4. ✕A 120-person software company with no job posted at allPartial fit and no timing signal: it stays on the list, lower down, until something changes.

Account scoring vs lead scoring: what is the difference?

Account scoring and lead scoring work together. The account score decides which companies to work; the lead score decides which person inside them to call first. A hot lead at a company with no fit is still a poor use of a rep’s day.

	Lead scoring	Account scoring
Scores	A person: a lead or a contact	A company
Built from	Title, seniority, and actions such as email clicks, visits and form fills	Fit with the ICP, and signals from the company and from all its people
Answers	Is this person ready for a sales conversation?	Is this company worth working now?
Typical use	Routing inbound leads to sales	Account-based marketing, territory planning, outbound lists

How do you build an account scoring model?

To build an account scoring model, start with fit: write down your ICP criteria and give the core ones the most weight. Add one or two timing signals you can count the same way for every company, then test the model on accounts you have already won: if your strongest customers would not have scored high, change the weights before you trust the order.

With BlitzAPI, Company Search returns the fit fields for the companies in your ICP, and Jobs Signal counts each company’s job postings by title and date, the timing signal used above. The build a TAM use case chains both into one list you can score.